Language Selection

English French German Italian Portuguese Spanish

Linux: Tuning CFS

Filed under
Linux

Nick Piggin used 'git bisect' to track a lmbench regression to the main CFS commit, leading to an interesting discussion between Nick and Ingo Molnar. Ultimately the regression was tracked down to the temporary configurability of the scheduler while it is tuned for optimal performance, "one reason for the extra overhead is the current tunability of CFS, but that is not fundamental, it's caused by the many knobs that CFS has at the moment." The solution, already coded but not yet merged in the mainline kernel "changes those knobs to constants, allowing the compiler to optimize the math better and reduce code size," and as a result result, "CFS can be faster at micro-context-switching than 2.6.22."

Ingo described the lmbench configuration in question as a "micro-benchmark", and noted that with a macro-benchmark better performance was more pronounced, "because with CFS the _quality_ of scheduling decisions has increased.

More Here




More in Tux Machines

Type Title Author Replies Last Postsort icon
Story Ubuntu's Mark Shuttleworth pulls no punches on Red Hat and VMware in OpenStack cloud Rianne Schestowitz 1 23/05/2018 - 1:06am
Story Security Bugs and FUD Roy Schestowitz 1 22/05/2018 - 11:23pm
Story Android Leftovers Rianne Schestowitz 22/05/2018 - 10:16pm
Story The Top 10 Endless Runner Games Mohd Sohail 22/05/2018 - 7:53pm
Story Here Is What's New In Fedora 28 Mohd Sohail 22/05/2018 - 7:48pm
Story Stable kernels 4.16.11, 4.14.43 and 4.9.102 Rianne Schestowitz 22/05/2018 - 7:18pm
Story today's leftovers Roy Schestowitz 22/05/2018 - 7:05pm
Story Software: Grafana, Heaptrack, Vim Roy Schestowitz 22/05/2018 - 7:04pm
Story GNU/Linux: Parrot 4.0, Oregan, Containers and Linux 4.18 Plans Roy Schestowitz 22/05/2018 - 7:01pm
Story Red Hat and Fedora Leftovers Roy Schestowitz 22/05/2018 - 5:44pm